Proceedings of the Polytechnic Association. 875 



it is only possible to cool an extremely thin sheet in this manner with 

 sufficient uniformity not to harden it. 



A thin plate of steel, such as described, can be hardened quite as 

 well with the hammer as in water ; saw plates are so haj-dened. A 

 planishing anvil is used, and a smooth-faced hammer, slightly rounded ; 

 the steel is condensed and strained by stretching unifonnly, by which 

 the tensions are inaugurated, giving the quality we call tempering. 



I suppose hundreds of mechanics have wondered, as I have, at these 

 mysterious changes in the qualities of steel by the process of harden- 

 ing, condensation and annealing, and have resorted to books, written 

 by learned men, for information, without finding their thirst for 

 knowledge gratified. 



It is difficult (probably impossible in cold water) to temper a large 

 piece of steel without cracking it. When plunged into the water the 

 outside is cooled suddenly and contracted cracks occur, which open 

 slightly at first, but afterward usuall}^ close, when the interior metal 

 cools and contracts, when the most dangerous ruptures occur. These 

 last cracks are only seen afterward, when the piece is entirely broken. 



A workman was wounded severely in Chicago a few years ago by 

 the explosion of a large steel step for a mill spindle, which he was 

 polishing in a lathe with emery ; the heat evolved by the friction 

 expanded the surface, which was already in a state of tension, from 

 the unequal cooling, but not quite ruptured. Heating the outside 

 under such circumstances was like " the feather which broke the 

 camel's back." 



In tempering large pieces of steel it is best not to make them 

 harder than required. Do not make them too hot, or have the water 

 too cold, especially if the temper is afterward to be drawn down ; 

 cracking the steel may in this manner be sometimes avoided, or the 

 piece may be cooled in oil, which will not cool it so rapidly, and 

 therefore will not cause so severe a tension. When it is practicable, 

 it is a good plan to make the piece hollow, by drilling a hole through 

 it by which the cooling surface will be increased, and greater hard- 

 ness obtained, wliile the rupturing tension will be less. 



If a flat, square, or round bar, of considerable length, is to be tem- 

 pered, hold it perpendicular as you insert it into the water ; then you 

 will not warp it. If you incline it to a considerable angle, it will 

 have the form of a segment of a circle ; the lower side will be long- 

 est. The part which cools last is always most reduced in length ; if 

 • any doubt this, let them try it before they tell anybody of their doubt. 



